A New Hope For Patients with Mild Gallstone Pancreatitis

Transcription

1 RADOMIZED CONTROLLED TRIALS Early Cholecystectomy Safely Decreases Hospital Stay in Patients With Mild Gallstone Pancreatitis A Randomized Prospective Study Armen Aboulian, MD,* Tony Chan, MD,* Arezou Yaghoubian, MD,* Amy H. Kaji, MD, PhD, Brant Putnam, MD,* Angela Neville, MD,* Bruce E. Stabile, MD,* and Christian de Virgilio, MD* Objective: We hypothesized that laparoscopic cholecystectomy performed within 48 hours of admission for mild gallstone pancreatitis, regardless of resolution of abdominal pain or abnormal laboratory values, would result in a shorter hospital stay. Summary of Background Data: Although there is consensus among surgeons that patients with gallstone pancreatitis should undergo cholecystectomy to prevent recurrence, the precise timing of laparoscopic cholecystectomy for mild disease remains controversial. Methods: Consecutive patients with mild pancreatitis (Ranson score 3) were prospectively randomized to either an early laparoscopic cholecystectomy group (within 48 hours of admission) versus a control laparoscopic cholecystectomy group (performed after resolution of abdominal pain and normalizing trend of laboratory enzymes). The primary end point was hospital length of stay. Secondary end point was a composite of rates of conversion to an open procedure, perioperative complications, and need for endoscopic retrograde cholangiography. The study was designed to enroll 100 patients with an interim analysis after 50 patients. Results: At interim analysis, 50 patients were enrolled at a single universityaffiliated public hospital. Of them, 25 patients were randomized to the early group and 25 patients to the control group. Patient age ranged from 18 to 74 years with a median duration of symptoms of 2 days upon presentation and a median Ranson score of 1. There were no baseline differences between the groups with regards to demographics, clinical presentation, or the presence of comorbidities. The hospital length of stay was shorter for the early cholecystectomy group (mean: % CI, , median: 3 IQR, 2 4 ) compared with the control group (mean: % CI, , median: 4 IQR, 4 6 P ). Six patients from the early group required endoscopic retrograde cholangiography, compared with 4 in the control group (P 0.72). There was no statistically significant difference in the need for conversion to an open procedure or in perioperative complication rates between the 2 groups. Conclusion: In mild gallstone pancreatitis, laparoscopic cholecystectomy performed within 48 hours of admission, regardless of the resolution of abdominal pain or laboratory abnormalities, results in a shorter hospital length of stay with no apparent impact on the technical difficulty of the procedure or perioperative complication rate. (Ann Surg 2010;251: ) From the *Departments of Surgery and Emergency Medicine, Harbor UCLA Medical Center, Torrance, CA; and Los Angeles Biomedical Research Intitute, Los Angeles, CA. Clinical investigators David Rosing, MD; Timothy Van Natta, MD; Stanley R. Klein, MD; and Frederic Bongard, MD. This study was registered at and carries the following ID NCT Reprints: Christian de Virgilio, MD, Department of Surgery, Harbor-UCLA Medical Center, Box 25, Torrance, CA cdevirgilio@ labiomed.org. Copyright 2010 by Lippincott Williams & Wilkins ISSN: /10/ DOI: /SLA.0b013e3181c38f1f Acute pancreatitis is a common diagnosis worldwide, with more than 220,000 cases reported annually in the United States alone. The leading etiology is gallstones. 1 Gallstone pancreatitis is considered to occur due to transient obstruction of the common channel that drains both the biliary and pancreatic ducts, resulting in inflammation of the pancreas. The pancreatitis that ensues is usually mild and self-limited and the treatment is initially supportive with subsequent laparoscopic cholecystectomy (LC). However, a small subgroup of patients develop severe pancreatitis and/or concomitant cholangitis. When the latter is present, ERC and sphincterotomy with stone extraction is typically performed. While there is a clear consensus that patients who present with gallstone pancreatitis should undergo cholecystectomy to prevent recurrence, precise timing of surgery remains controversial. In patients with severe pancreatitis (Ranson score 3), there is consensus that surgery is delayed until the pancreatitis has resolved because early operation is associated with a higher complication rate. 2 However, despite more than 30 years of debate in the surgical literature, the optimal timing of surgery in mild pancreatitis (Ranson score 3) remains unclear. With recurrence rates for gallstone pancreatitis reported as high as 63% 3 and with some of the repeat attacks occurring within 2 weeks of initial index presentation, 1 most investigators have recommended cholecystectomy during the initial hospitalization. 4,5 Still, the actual timing of surgery during the initial index hospitalization is unsettled. In practice, surgeons often delay surgery until there is evidence of complete resolution of the inflammatory process, as evidenced by absence of abdominal pain and normalization of liver functional tests and pancreatic enzymes. 6 Unfortunately, this strategy may result in prolongation of hospitalization without any proven benefit. A previous prospective, nonrandomized study from our institution suggested that early cholecystectomy could safely be performed within 48 hours of admission in patients with mild pancreatitis, regardless of resolution of abdominal pain and abnormal laboratory values. In this study, when compared with a retrospective control group in which surgery was delayed until there was resolution of clinical and laboratory parameters, hospital stay was significantly reduced from a median of 7 to 4 days, without additional complications. 7 To address the optimal timing of surgery, the results and overall outcomes of a prospective randomized study are presented here in which patients with mild gallstone pancreatitis were allocated to either an early group (surgery within 48 hours of presentation) or a control group (surgery after resolution of abdominal pain and normalization of laboratory values). PATIENTS AND METHODS Study Setting The study was performed at Harbor-UCLA Medical Center, a Los Angeles County public teaching hospital in Torrance, California. It is a level I trauma center and serves a population over 3 million people in the Los Angeles area. Annals of Surgery Volume 251, Number 4, April

2 Aboulian et al Annals of Surgery Volume 251, Number 4, April 2010 Hypothesis In patients with mild pancreatitis, early cholecystectomy (within 48 hours of admission and potentially before normalization of laboratory values and resolution of abdominal symptoms) can safely be performed without additional complications and results in a shorter hospital stay as compared with delaying cholecystectomy until laboratory values and abdominal pain resolve. Inclusion Criteria All adults between the age of 18 and 100 with mild gallstone pancreatitis were included. A subject was classified as having gallstone pancreatitis if they had the following: (a) upper abdominal pain, nausea, vomiting, and epigastric tenderness; (b) absence of ethanol abuse; (c) elevated amylase level to at least twice the upper limit of normal and elevated lipase level to at lease 3 times the upper limit of normal; and (d) imaging confirmation of gallstones. The classification of mild pancreatitis was defined by the presence of the following: (a) 3 or fewer Ranson criteria on admission: age 55 years, glucose 200 mg/dl, LDH 350 mg/dl, AST 250 units/l, and WBC 16 K/mm 3 ;(b) clinical stability with admission to a nonmonitored ward bed; (c) absence of acute cholangitis defined as a temperature 38.6 C, right upper quadrant pain and tenderness, and significant hyperbilirubinemia; and (d) low suspicion for a retained common bile duct (CBD) stone (total bilirubin 4 mg/dl on admission). Exclusion Criteria Exclusion criteria were as follows: (a) severe pancreatitis (as defined by the presence of more than 3 Ranson criteria on admission); (b) suspected concomitant acute cholangitis; (c) high suspicion for retained CBD stone (total bilirubin 4 mg/dl on admission or ultrasound demonstration of CBD stone); (d) patient refusal to participate; (e) severe preexisting medical comorbidities contraindicating cholecystectomy (as determined by the primary physicians); (f) pregnancy; (g) prior gastric bypass surgery (making ERC difficult); (h) admission to a monitored unit. The need for admission to a monitored unit was determined by the admitting surgeon and was guided primarily by a need for aggressive fluid administration as demonstrated by severe volume depletion (eg, on admission tachycardia 110 beats/min, blood urea nitrogen 15 mg/dl) or evidence of cholangitis. Study Design and Randomization Patients deemed to have mild gallstone pancreatitis that met the inclusion criteria and gave informed consent were prospectively randomized to an early group versus a control group. Random assignment was performed by drawing a sealed, unlabeled, unordered envelope from a container by an independent party immediately after informed consent was obtained. In the patients randomized to the early group, LC with intraoperative cholangiography (IOC) was performed within 48 hours of admission, regardless of whether or not abdominal pain and tenderness were still present and laboratory values had normalized. In the control group, LC with IOC was performed only after resolution of abdominal pain and normalization of laboratory values. The study was designed to have an interim analysis after the enrollment of 50 patients to assess for differences in outcome. The study was approved by the institutional review board and the Committee on Human Subjects at the Los Angeles Biomedical Institute of Harbor-UCLA Medical Center. Primary Endpoints The primary end point for the study was length of hospital stay. Secondary Endpoints Secondary end point was a composite of (1) need for conversion to an open cholecystectomy, (2) need for ERC, and (3) perioperative complications including bile duct injury, bleeding requiring transfusion or reoperation, wound infection, pneumonia, and need for readmission within 30 days. Surgery and Postoperative Management All operations were performed by general surgery residents under the supervision of the faculty surgeon. All patients received appropriate perioperative antibiotics. All cholecystectomy surgeries were performed laparoscopically with a standard 4 port technique. Special modifications were made to the procedure as deemed necessary by the surgical team. These included direct cut-down (Hassan) technique when placing the first subumbilical port, use of a 10 mm or an angled scope, decompression of the gallbladder, suture closure of the cystic duct, and use of an endoscopic bag to retrieve the gallbladder. The gallbladder was dissected off the liver using monopolar electrocautery. An IOC was attempted on every patient. Patients were transferred back to a ward setting after the surgery. They received a liquid diet for the first meal and subsequently advanced to a regular diet. Patients were discharged on postoperative day 1 if they were able to tolerate their diet, their pain was adequately controlled with oral medication, and they had no other indications for continued hospitalization. Data Collection Preoperative, operative, and postoperative data were prospectively collected for each patient participating in the study. Preoperative data included demographics, medical history, and examination findings and routine laboratory values and abdominal ultrasound results. Operative data included time from admission to operation and intraoperative complications and IOC results. Postoperative data included hospital course and complications, need for ERC with respective findings, overall length of stay, and readmission within 30 days. Data Analysis For the purpose of study planning, the sample size requirements to demonstrate an association between the intervention and length of hospital stay were calculated using PASS 2002 Software (NCSS Corporation, Kaysville, UT). The study was powered for length of stay. A sample size of 100 patients was chosen to achieve 90% power to detect a mean 2 day (SD 1 day) difference in the early group versus the control group. A significance level (alpha) of 0.05 was chosen with an interim analysis planned after 50 patients. Using the O Brien-Fleming method, it was determined that the study could be terminated after 50 patients if there was a difference in length of hospital stay between the 2 groups with a predefined and more stringent alpha level of If this alpha level was not achieved after enrolling 50 patients, according to the O Brien- Fleming method, an alpha level of would have to be achieved at the final analysis to demonstrate a statistically significant difference between the 2 groups. 8 Statistical analysis was performed using SAS version 9.2 (SAS Institute, Cary, NC). Results were expressed as means with 95% confidence intervals (CI) or medians with interquartile ranges (IQRs) for continuous variables, and the nonparametric Wilcoxon rank sum or Kruskal-Wallis tests were performed to assess for significant differences between the 2 groups. For categorical variables, the 2 or Fisher exact test tests were performed, as appropriate, to compare baseline demographic variables, duration of symptoms, vital signs, laboratory data, use of ERC, and rates of other complications. Analysis of variance was used to compare the 2 cohorts with regards to the baseline categorical variable of race Lippincott Williams & Wilkins

3 Annals of Surgery Volume 251, Number 4, April 2010 Early Cholecystectomy Decreases Hospital Stay RESULTS Patient Recruitment and Study Progress From November 2007 to November 2008, 84 patients presented with a primary diagnosis of gallstone pancreatitis. Of the 84 patients, 23 did not meet inclusion criteria based on the following 14 patients had a high suspicion for a retained CBD stone, 6 patients were admitted to a monitored bed for severe pancreatitis, cholangitis, or need for aggressive resuscitation. One patient had a history of pancreatitis and had previously undergone an endoscopic sphincterotomy, 1 patient had undergone a gastric bypass which would make ERC difficult if needed, and 1 patient was deemed medically unfit to undergo surgery. The remaining 61 patients qualified for the study according to the inclusion and exclusion criteria; however, 10 patients refused to participate in the study. One patient was excluded after randomization when it was determined that the cause of the pancreatitis was severe hypertriglyceridemia. All enrolled patients but one underwent LC. This one patient, who randomized to the control group, developed respiratory failure and myocardial infarction prior to any procedures and required long term intensive care unit hospitalization and never underwent an operation. The data for this patient were excluded from analysis as they were incomplete. Patient Characteristics and Admission Values The analysis included 25 patients in the early group and 24 in the control group. Patient characteristics and admission laboratory values are shown in Table 1. There were no differences with respect to median age, gender, duration of symptoms, diabetes incidence, admission vital signs, and Bedside Index for Severity in Acute Pancreatitis score, a useful scoring system in evaluating the severity of pancreatitis via patient age 60, pleural effusion, symptoms of SIRS, GCS 15, and BUN There was a statistically but not clinically significant difference in admission serum creatinine in the early group (0.7 mg/dl) versus the control group (0.6 mg/dl) (P 0.04).There were no differences in other admission laboratory values. The median admission Ranson score was 1 in both groups (P 0.74). The mean CBD size on admission ultrasound was similar in the early group (5.9 mm) versus the control group (6.7 mm) (P 0.68). Patient Management All patients in the study had a second set of laboratory tests and clinical reassessment to ensure improvement or at least stability of the pancreatitis. Patients in the early group were taken to the operating room at a mean of 35.1 hours after admission compared with 77.8 hours in the control group (P 0.001). One patient in the early group was taken to the operating room more than 48 hours from time of admission due to unavailability of the operating room. All cholecystectomies were successfully completed laparoscopically without conversion to open. IOC was unable to be obtained due to technical difficulties in 1 (4%) patient in the early group and in 4 patients (16.6%) in the control group (P 0.19). CBD stones or lack of duodenal filling were found on IOC in 6 patients in the early group and 3 of the patients in the control group (P 0.27). These patients underwent a successful postoperative ERC and stone retrieval. There was 1 patient in the control arm who underwent an ERC prior to surgery for increase in bilirubin. Primary Endpoint The overall length of hospital stay was shorter for the early cholecystectomy group (mean: % CI, , median: 3 IQR, 2 4 ) compared with the control group (mean: % CI, , median: 4 IQR, 4 6, P ). TABLE 1. Patient Characteristics and Admission Values Early Group (n 25) Median (IQR) Control Group (n 25) Median (IQR) Age (yr) 33 (29 51) 41 (30 50) 0.71 Gender 23 men; 2 women 22 men; 3 women 0.60 Comorbidities (No. patients) Duration of symptoms (d) 2 (1 6) 2 (1 3) 0.94 Heart rate 74 (68 80) 73 (67 84) 0.85 Systolic blood pressure (mm Hg) 123 ( ) 121 ( ) 0.48 Diastolic blood pressure (mm Hg) 70 (67 87) 70 (67 77) 0.44 Temperature (F) 97.6 ( ) 97.9 ( ) 0.38 WBC (K/mm 3 ) 11.0 ( ) 10.7 ( ) 0.92 Hemoglobin (g/dl) 13.5 ( ) 13.2 ( ) 0.83 Hematocrit (%) 38.7 ( ) 38.8 ( ) 0.75 Creatinine (mg/dl) 0.7 ( ) 0.7 ( ) 0.04 Total bilirubin (mg/dl) 1.60 ( ) 2 (1 2.6) 0.41 ALT (units/l) 225 (79 432) 306 ( ) 0.40 AST (units/l) 157 (81 323) (25 597) 0.47 Alkaline phosphatase (units/l) 146 (99 245) 195 ( ) 0.66 LDH (mg/dl) ( ) 202 ( ) 0.15 Amylase (mg/dl) 486 ( ) 1204 ( ) 0.07 Lipase (mg/dl) 800 ( ) 1888 ( ) 0.06 CBD size on ultrasound (mm) 5.2 ( ) 5.8 (4 8) 0.68 BISAP score (mean with 95% CI) 0.4 (0.08, 0.72) 0.29 (0.08, 0.52) 0.77 Ranson score 1 (0 1) 1 (0 1) 0.74 WBC indicates white blood cell; ALT, alanine transaminase; AST, aspartate transaminase; LDH, lactate dehydrogenase; CBD, common bile duct; IQR, interquartile range; CI, confidence interval. P 2010 Lippincott Williams & Wilkins 617

4 Aboulian et al Annals of Surgery Volume 251, Number 4, April 2010 TABLE 2. Primary and Secondary Endpoints Early Group Control Group P/Odds Ratio Primary endpoint Length of stay (d) 3 (2 4) 4 (4 6) Secondary endpoint Overall rate /1.66 Conversion rate 0 0 ERC /1.66 Complication rate ERC indicates endoscopic retrograde cholangiography. Secondary Endpoints There were no patients in either group that required conversion to open cholecystectomy and no bleeding requiring transfusion, postoperative complications, or readmissions. In the early group, there were 6 secondary endpoints in 6 patients (24%) versus 4 secondary endpoints in 4 patients (17%) in the control group. The secondary endpoints identified in patients in both groups of the study were the need for ERC (prompted by either a CBD stone seen on IOC or for failure to visualize contrast in the duodenum). This difference was not statistically significant (P 0.48, OR: 1.66, 95% CI: ) Table 2. Interim Analysis As planned, an interim analysis was performed after 50 patients were enrolled. The respective mean and median lengths of stay for the 2 groups were 3.5 and 3 days in the early group (IQR 3 6) versus 5.8 and 4 days (IQR 4 6) in the control group (P ). This P value was notably less than the more stringent predefined P the cut-off value calculated a priori, necessary to terminate patient enrollment. The difference between the 2 groups remained statistically significant even after adjusting for covariates and after performing sensitivity analysis to remove outliers. Furthermore, a significant difference in hospital stay was noted without a significant increase in the secondary endpoints: rates of conversion to open cholecystectomy, complications rates, and need for pre- or postoperative ERC. The results of the interim analysis indicated early termination of the study would be appropriate, and hence, patient enrollment was terminated. DISCUSSION In a randomized prospective study of patients with mild gallstone pancreatitis, the findings of this study demonstrate that LC with IOC performed within 48 hours of admission irrespective of normalization of laboratory values or clinical progress, decreases the overall length of hospital stay from 4 to 3 days compared with delaying LC until laboratory values and clinical condition normalize. In addition, there were no differences in the secondary endpoints of need for conversion to open operation, need for ERC, or complication rates. The study was terminated at the interim analysis as the difference reached a more stringent value of significance (P 0.005). The timing of cholecystectomy in patients with mild gallstone pancreatitis has been controversial. In the era of open cholecystectomy, Kelly and Wagner performed a randomized study to answer this question. Patients were classified into 2 risk groups based on Ranson prognostic signs of severity: mild pancreatitis ( 3 Ranson signs) and severe pancreatitis ( 3 Ranson signs). In patients with severe pancreatitis, early (within 48 hours) open cholecystectomy was associated with a morbidity and mortality of 82.6% and 47.8%, respectively, versus 17.6% morbidity and 11% mortality with delayed surgery (P 0.001). As such, delayed cholecystectomy in these patients was recommended. However, in patients with mild pancreatitis, there was no difference in morbidity and mortality. 2 Despite the findings of the aforementioned study, the precise timing of surgery in the laparoscopic era in patients with mild pancreatitis remains a topic of debate, with a tendency to adhere to traditional methods of awaiting resolution of abdominal pain and normalization of laboratory values. In a recent editorial, Vitale recommended performing delayed cholecystectomy only after a resolving enzyme pattern and decreasing bilirubin are documented. 10 In another recent study, patients were hospitalized until laboratory values and physical examination normalized, subsequently discharged home without cholecystectomy, and then scheduled for an elective cholecystectomy as an outpatient within 2 weeks after discharge. 11 Patients were hospitalized a mean of 5.9 days at the initial hospitalization, and the mean time to cholecystectomy using their pathway was 22.7 days. In our opinion, this approach creates unnecessary delays in surgical therapy and a potential for recurrent pancreatitis. This concern was validated by Ito et al, who reported that 32% of discharged patients were readmitted for recurrent pancreatitis prior to their scheduled elective cholecystectomy, and of those, nearly a third recurred within 2 weeks following discharge. 1 This approach is contrasted with the study being presented here, in which patients undergo a single admission and have their cholecystectomy with a mean hospital length of stay of only 3.5 days. Other authors have advocated a more aggressive approach towards early cholecystectomy in patients with mild gallstone pancreatitis. In a retrospective study, Taylor and Wong compared the approach of 2 surgeons for mild pancreatitis. The first surgeon delayed surgery until normalization of amylase and complete resolution of abdominal tenderness, whereas the second surgeon proceeded with LC as soon as the amylase was decreasing and the abdominal pain was improving. There were no differences in the complication rates which were 10% and 11%, respectively (P 0.12). However, early cholecystectomy was associated with a significantly shorter hospital stay, 3.5 days as compared with 4.7 days in the delayed surgery group (P 0.01). 12 This approach was verified by Rosing et al who instituted a practice of early cholecystectomy (within 48 hours) in a prospective study of 43 patients. The median length of hospital stay was 4 days, as compared with a 7-day hospital stay in a retrospective group of 177 patients treated with delayed cholecystectomy, before implementation of this policy. Complication rates were 4.8% (early cholecystectomy group) and 4.5% (delayed cholecystectomy group) (P 0.7). 6 The conclusions of these studies, that earlier surgery results in a shorter hospital stay without an increase in complication rates, are consistent with the present study. An important requirement for successful early cholecystectomy is appropriate patient selection, which is predicated on the fact that suitable candidates can be identified on admission or early in the hospitalization course. Contraindications to early cholecystectomy include the need for aggressive fluid resuscitation in a monitored setting, concomitant cholangitis, and severe pancreatitis. On the basis of a previous study by Arnell et al, 2 factors were identified as predictive of a need for aggressive fluid resuscitation in a monitored setting: tachycardia 110 beats per minute and a BUN 15 mg/ dl. 13 Prior studies indicate that concomitant cholangitis with gallstone pancreatitis is infrequent, and was noted in only 3% of patients in 1 prospective study. 14 We have observed that patients who have cholangitis almost always manifest signs and symptoms on admission, and very rarely develop cholangitis later in their hospital course. The findings of Charcot triad is not reliable in establishing the diagnosis of cholangitis, as it is present in less than half of patients with cholangitis. Rather, cholangitis should be suspected in Lippincott Williams & Wilkins

5 Annals of Surgery Volume 251, Number 4, April 2010 Early Cholecystectomy Decreases Hospital Stay patients with markedly elevated bilirubin ( 4 mg/dl) and alkaline phosphatase, fever, and leukocytosis. Such patients should undergo urgent ERC. Identifying patients with severe pancreatitis on admission is potentially hampered by the fact that Ranson criteria requires 48 hours to calculate. We defined severe pancreatitis as having 3 or more Ranson criteria on admission. Using this definition, only 1 patient (2%) in our study progressed to developing a severe complication (myocardial infarction and respiratory failure) that required ICU admission. Furthermore, all of the patients in the early arm of our study had a second set of laboratory tests and clinical reassessment prior to surgery. This approach allowed for documentation of improvement or at least stability of the pancreatitis. Although the average time from admission to surgery was 35.1 hours in the early arm, there was 1 patient that was taken to the operating room within 6 hours of admission. However, even in that patient, improvement was confirmed by laboratory tests and clinical evaluation. In the setting of gallstone pancreatitis, about 85% of CBD stones pass spontaneously into the gastrointestinal tract, and most do so by 48 hours of admission. 15 Hence, a potential drawback of early cholecystectomy is that common duct stones may not be cleared prior to cholecystectomy and those that undergo early cholecystectomy may have higher rates of CBD stones on IOC and a greater need for ERC in the postoperative period. Although this study was not powered to identify a difference in CBD stones on IOC, there was not a significant difference in the requirement for ERC between the 2 groups. Although it would be tempting to perform preoperative ERC in everyone, so as to definitively clear the common duct, or wait for as many stones to pass as possible, such an approach would lead to unnecessary ERCs and extended days of hospitalization, and in the absence of cholangitis, is unnecessary. 16 In a randomized study of patients at high risk for a retained CBD stone, selective postoperative ERC resulted in 100% duct clearance, and was found to shorten the overall average length of stay and lower cost when compared with routine preoperative ERC. Total bilirubin appears to be the best predictor of a persisting CBD stone in gallstone pancreatitis. 17 A total bilirubin 4 mg/dl on admission or on hospital day 2 has the best positive predictive value, 18 and avoids unnecessary preoperative ERC. Patients who met this bilirubin threshold were excluded from the study. We do not feel that a bilirubin 4 mg/dl is a contraindication to early cholecystectomy per se. However, since these patients have a much higher chance of needing an ERC with a resulting longer hospitalization, such patients may potentially confound our primary end point of hospital stay. In addition, with the bilirubin above this threshold, the likelihood of cholangitis significantly increases. It should be noted that there are several potential limitations of our study. The study is a single center experience with a relatively small sample size. However, the findings are supported by prior studies. Although there was no significant difference in our secondary end point (need for conversion to open cholecystectomy, need for ERC, and complications), the study design was powered to detect a difference in length of stay and not sufficiently powered to detect a difference in the secondary end point (95% CI, ). To perform a study specifically powered for differences in complication rates, as majority of patients with mild gallstone pancreatitis have an extremely low rate of complications, a significantly higher number of patients would need to be enrolled. Furthermore, the majority of our patients were young and Hispanic, reflecting our indigent public hospital population. Prior studies have shown that Hispanic patients with gallstone pancreatitis have a relatively benign course. 19 However, since our study specifically focused on mild pancreatitis, our results support the premise that there is no need to delay cholecystectomy until enzymes and physical examination normalize. In conclusion, this prospective, randomized study indicates that LC can be safely performed within 48 hours of admission in patients with mild gallstone pancreatitis. As further confirmed by our study (61 of 84 patients 73% were eligible for enrollment), most patients with gallstone pancreatitis have a milder form of pancreatitis and may undergo early LC. The practice of delay until laboratory values and physical examination normalize should be abandoned as it results in a prolonged hospital stay. This approach is ideally suited to patients with mild gallstone pancreatitis who do not demonstrate evidence of cholangitis or the need for aggressive fluid resuscitation. REFERENCES 1. Ito K, Ito H, Whang EE. Timing of cholecystectomy for biliary pancreatitis: do the data support current guidelines? J Gastrointest Surg. 2008;12: Kelly TR, Wagner DS. Gallstone pancreatitis: a prospective randomized trial of the timing of surgery. Surgery. 1988;104: Uhl W, Muller CA, Krahenbuhl L, et al. Acute gallstone pancreatitis. Surg Endosc. 1999;13: Kelly T, Elliot DW. The management of gallstone pancreatitis in the era of laparoscopic cholecystectomy. Am J Surg. 1996;172: Alimoglu O, Ozkan O, Sahin M, et al. Timing of cholecystectomy for acute biliary pancreatitis: outcomes of cholecystectomy on first admission and after recurrent biliary pancreatitis. World J Surg. 2003;27: Larson SD, Nealon WH, Evers BM. Management of gallstone pancreatitis. Adv Surg. 2006;40: Rosing DK, devirgilio C, Yaghoubian A, et al. Early cholecystectomy for mild to moderate gallstone pancreatitis shortens hospital stay. J Am Coll Surg. 2007;205: Lewis RJ. An introduction to the use of interim data analyses in clinical trials. Ann Emerg Med. 1993;22: Singh VK, Wu BU, Bollen TL, et al. A prospective evaluation of the bedside index for severity in acute pancreatitis score in assessing mortality and intermediate markers of severity in acute pancreatitis. Am J Gastroenterol. 2009;104: Vitale GC. Early management of acute gallstone pancreatitis. Ann Surg. 2007;245: Clarke T, Sohn H, Kelso R, et al. Planned early discharge-elective surgical readmission pathway for patients with gallstone pancreatitis. Arch Surg. 2008;143: Taylor E, Wong C. The optimal timing of laparoscopic cholecystectomy in mild gallstone pancreatitis. Am Surg. 2004;70: Arnell TD, devirgilio C, Chang L, et al. Admission factor can predict the need for ICU monitoring in gallstone pancreatitis. Am Surg. 1996;62: Chang L, Lo S, Stabile BE, et al. Preoperative versus postoperative endoscopic retrograde cholangiopancreatography in mild to moderate gallstone pancreatitis. Ann Surg. 2000;231: Acosta JM, Katkhouda N, Bebian KA, et al. Early ductal decompression versus conservative management for gallstone pancreatitis with ampullary obstruction. Ann Surg. 2006;243: Petrov M, van Santvoort HC, Besselink MGH, et al. Early endoscopic retrograde cholangiopancreatography versus conservative management in acute biliary pancreatitis without cholangitis. Ann Surg. 2008;247: Chang L, Lo SK, Stabile BE, et al. Gallstone pancreatitis: a prospective study on the incidence of cholangitis and clinical predictors of retained common bile duct stones. Am J Gastroenterol. 1998;93: Chan T, Yagoubian A, Rosing DK, et al. Total bilirubin is a useful predictor of persisting common bile duct stone in gallstone pancreatitis. Am Surg. 2008;74: Yaghoubian A, devirgilio C, El-Masry M, et al. Gallstone pancreatitis: a benign disease in Hispanics. Am Surg. 2007;73: Lippincott Williams & Wilkins 619